Research on yoga for chronic low back pain generally shows modest improvements in pain and back-related function, especially over the first 3 to 6 months. Results vary by comparison group, outcome scale, program length, and follow-up period. The figures below come from systematic reviews and randomized trials, with measurement periods and populations kept explicit.
Contents
- Pain outcomes
- Disability and physical function
- Quality of life, sleep, and mental health
- Yoga compared with exercise and physical therapy
- Participation, durability, and medication use
- Clinical improvement and safety
Pain outcomes
The size of the reported pain benefit depends partly on the measurement scale and the control group. A Cochrane review, with evidence searched through 31 August 2021, included 9 trials and found that yoga reduced pain by 4.53 points on a 0–100 scale versus no exercise at 3 months. The 95% confidence interval was 2.46 to 6.61 points in favor of yoga, and the analysis included 946 participants. The review rated this pooled pain evidence as moderate certainty (Cochrane review, Yoga for chronic non-specific low back pain).
An 18-trial meta-analysis searched through 1 November 2019 and reported smaller standardized-scale differences versus non-exercise controls. The pain reduction was 0.83 points at 4–8 weeks, 0.43 points at 3 months, and 0.56 points at 6–7 months. Their 95% confidence intervals were 0.48–1.19, 0.23–0.64, and 0.11–1.02 points, respectively, in favor of yoga. Statistical heterogeneity was 0% at 4–8 weeks and 3 months, and 50% at 6–7 months. At 12 months, the pooled difference was 0.52 points in favor of yoga, but the confidence interval crossed no difference: 0.59 points against to 1.64 points in favor, with I²=87% (systematic review and meta-analysis).
Several trials provide results in more familiar clinical terms. A 12-week meta-analysis of six trials estimated a within-trial pain effect of SMD −0.41, with a 95% confidence interval from −0.58 to −0.23 and p<0.0001. In a 2009 pilot trial conducted predominantly among minority participants, mean pain fell 2.3 points in the yoga group and 0.4 points in controls over 12 weeks, p=0.02. A prespecified clinically significant pain decrease was reached by 67% of yoga participants versus 13% of controls; the odds ratio was 5.0, with a 95% confidence interval of 1.3–19.1 (12-week meta-analysis; pilot trial).
In a virtual-yoga randomized clinical trial reported in 2024, pain intensity was 1.5 points lower with yoga than delayed yoga at week 12 on an 11-point scale. The 95% confidence interval was 0.7–2.2 points, p<0.001. The trial began with 140 adults whose mean baseline pain intensity was 5.7/10 (virtual-yoga trial).
Disability and physical function
Back-specific disability outcomes are often measured with the Roland-Morris Disability Questionnaire, or RMDQ. It ranges from 0 to 24 in the Cochrane analysis. At 3 months, yoga improved RMDQ function by 1.69 points versus no exercise, with a 95% confidence interval of 0.65–2.73 points. The Cochrane review rated this pooled back-function evidence as low certainty (Cochrane review).
The 18-trial meta-analysis found disability benefits versus non-exercise controls at several time points:
| Follow-up | Disability result | Heterogeneity |
|---|---|---|
| 4–8 weeks | SMD −0.30 (95% CI −0.51 to −0.10) | I²=0% |
| 3 months | SMD −0.31 (95% CI −0.45 to −0.18) | I²=30% |
| 6 months | SMD −0.38 (95% CI −0.53 to −0.23) | I²=0% |
| 12 months | SMD −0.33 (95% CI −0.54 to −0.12) | I²=9% |
These estimates were reported in the 2019 evidence search and compare yoga with non-exercise controls (systematic review and meta-analysis).
A UK randomized trial conducted from April 2007 through March 2010 reported adjusted RMDQ differences favoring yoga over usual care of 2.17 points at 3 months, 1.48 points at 6 months, and 1.57 points at 12 months. The corresponding 95% confidence intervals were 1.03–3.31, 0.33–2.62, and 0.42–2.71 points (UK randomized trial).
The 2024 virtual-yoga trial found a 2.8-point greater RMDQ reduction than delayed yoga at week 12, with a 95% confidence interval of 1.3–4.3 points and p<0.001. By week 24, the yoga group’s mean RMDQ change was −4.6 points, with a 95% confidence interval from −6.1 to −3.1 and p<0.001 (virtual-yoga trial).
Quality of life, sleep, and mental health
At about 3 months, the Cochrane review found that yoga improved SF-36 physical health scores by 1.80 points versus no exercise. The 95% confidence interval was 0.27–3.33 points, based on 686 participants. SF-36 mental health scores improved by 2.38 points, with a 95% confidence interval of 0.60–4.17 points, also among 686 participants (Cochrane review).
A 27-study review, with evidence searched through 26 May 2020, found short-term mental-health improvement versus passive controls of 1.70 points across 7 randomized controlled trials; the 95% confidence interval was 0.20–3.20. Expressed as a standardized effect, the estimate was SMD 0.17, with a 95% confidence interval of 0.02–0.32. Short-term physical functioning improved by 2.80 points across 9 randomized controlled trials, with a 95% confidence interval of 1.00–4.70; the standardized estimate was SMD 0.28, with a 95% confidence interval of 0.10–0.47 (systematic review and meta-analysis).
Flexibility-related results were reported in a one-week residential Indian trial published in 2010. Right-leg straight-leg raising improved 31.1% with yoga versus 18.7% with physical therapy. Left-leg straight-leg raising improved 28.4% with yoga versus 21.5% with physical therapy (Indian randomized study).
Sleep also improved in the virtual-yoga trial. At week 12, the sleep-quality score was 0.4 points better with yoga than delayed yoga, with a 95% confidence interval of 0.1–0.7 and p=0.008. The advantage remained 0.4 points at week 24, with the same confidence interval and p=0.005 (virtual-yoga trial).
Yoga compared with exercise and physical therapy
Compared with no exercise, the Cochrane review estimated that participants receiving yoga were more likely to report improvement or resolution of back pain at about 3 months. The risk ratio was 2.33, with a 95% confidence interval of 1.46–3.71 across 4 trials (Cochrane review).
Yoga and physical therapy produced broadly similar results in a 2017 noninferiority trial. At 12 weeks, mean RMDQ improvement was −3.8 points with yoga and −3.5 points with physical therapy. Mean pain improvement was −1.7 points with yoga and −2.3 points with physical therapy on an 11-point scale. The one-sided 95% lower confidence limit for yoga versus physical therapy was 0.83 RMDQ points, within the prespecified 1.5-point margin, and 0.97 pain points, within the 1.0-point margin (noninferiority trial).
The same trial found that yoga participants were 21 percentage points less likely than education participants to use pain medication at 12 weeks. Physical-therapy participants were 22 percentage points less likely than education participants to use pain medication. Improvements in the yoga and physical-therapy groups were maintained at one year regardless of maintenance strategy.
A 2026 meta-analysis, based on an evidence search through 23 November 2025, found no statistically significant pain advantage for yoga over exercise: SMD −0.52, 95% CI −1.38 to 0.35, I²=95%. Based on 2 studies, yoga improved physical function versus exercise with SMD −1.20, 95% CI −1.64 to −0.77, I²=31%. The same synthesis found no clear disability difference, SMD −0.19, 95% CI −1.11 to 0.73, I²=92%, but better emotional wellbeing with yoga, SMD −0.71, 95% CI −1.26 to −0.16, I²=75% (2026 systematic review and meta-analysis).
Participation, durability, and medication use
Program structure and attendance matter when interpreting trial results. In the UK trial, the intervention offered 12 gradually progressive yoga classes across 3 months. Of 156 people offered yoga, 93, or 60%, attended at least 3 of the first 6 sessions and at least 3 additional sessions (UK randomized trial).
The veteran randomized trial enrolled participants between 2013 and 2015; mean back-pain duration at enrollment was 15 years. Yoga was delivered twice weekly for 12 weeks with home practice. At 6 months, veterans assigned to yoga had a 2.48-point greater RMDQ reduction than those assigned to delayed treatment, with a 95% confidence interval of 0.87–4.08 points (veteran trial).
Medication-use differences were also reported in the virtual-yoga trial. Any analgesic use was 21.4 percentage points lower with yoga than delayed yoga at week 12, with a 95% confidence interval of 5.2–37.6 points. At week 24, the reduction was 21.2 percentage points, with a 95% confidence interval of 5.2–37.3 points (virtual-yoga trial).
Clinical improvement and safety
Safety findings were generally reassuring but not uniform. In the UK trial, adverse events were reported by 12 of 156 yoga participants versus 2 of 157 usual-care participants; most were increased pain. In the predominantly minority 12-week pilot trial, one yoga participant reported transient worsening of low back pain, and no other significant adverse event was reported (UK randomized trial; pilot trial).
When yoga was compared with other exercise, the Cochrane review found a similar adverse-event risk at 6–12 months: risk ratio 0.93, with a 95% confidence interval of 0.56–1.53. The corresponding figures were 84 events per 1,000 participants with yoga and 91 per 1,000 with other exercise (Cochrane review). No serious adverse events were reported during the 12-week intervention or follow-up in the veteran yoga trial (veteran trial).
The evidence base still has important limits. The Cochrane review reported 0 randomized trials comparing yoga with sham yoga. Its pooled 3-month pain evidence was rated moderate certainty, while pooled 3-month back-function evidence was rated low certainty. The 2026 comparison with exercise also showed substantial heterogeneity for pain, disability, and emotional wellbeing, so pooled estimates should not be treated as a guarantee for an individual program or participant (Cochrane review; 2026 systematic review and meta-analysis).